From: John W Higgins <wishdev@...>
Date: 2010-11-12T08:14:05+09:00
Subject: Re: WEBrick HTTPProxyServer question

--000e0cd47c482fadda0494cf21be
Content-Type: text/plain; charset=ISO-8859-1

Good Afternoon

On Thu, Nov 11, 2010 at 11:06 AM, Nokan Emiro <uzleepito@gmail.com> wrote:

>   Hi,
>
> could somebody please show me what can I do
> to prevent WEBrick::HTTPProxyServer from
> accessing the webserver on special cases?  There
> are cases when the proxy already knows in the
> RequestCallback procedure that the content on
> the webserver is irrelevant to the data sent back
> to the browser.  Think about a http proxy, that
> has a list of URLs that should be blocked, or
> think about a caching proxy, that does not need
> to contact the webserver if the content is already
> available in the RequestCallback phase:
>
> require 'webrick/httpproxy'
> server = WEBrick::HTTPProxyServer.new(
>  :Port => 8080,
>  :RequestCallback => Proc.new do |req, res|
>    if ## condition ##
>        ## what to do here to prevent http req? ##
>    end
>  end
> )
> server.start
>
>
First, it appears that RequestCallback merely allows for the modification of
the initial request, so you could modify the request to point to a cache
server and have it serve the initial request if you wanted.

The other option is to subclass the HTTPProxyServer class and do what you
want from within.

Quick and dirty concept
require 'webrick/httpproxy'

class X < WEBrick::HTTPProxyServer
  def do_GET(req, res) #there are also do_POST and friends that you might
want to override as well
     if req.request_uri = 'blah blah blah'
       #do whatever you want here
       #you will need to setup the res response object to properly look like
an HTTP Response
    else
       super
    end
  end
end

server = X.new(
 :Port => 8080
)
server.start


John

--000e0cd47c482fadda0494cf21be--